Training Feature Attribution for Vision Models
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912640537722880 |
|---|---|
| author | Bacha, Aziz George, Thomas |
| author_facet | Bacha, Aziz George, Thomas |
| contents | Deep neural networks are often considered opaque systems, prompting the need for explainability methods to improve trust and accountability. Existing approaches typically attribute test-time predictions either to input features (e.g., pixels in an image) or to influential training examples. We argue that both perspectives should be studied jointly. This work explores *training feature attribution*, which links test predictions to specific regions of specific training images and thereby provides new insights into the inner workings of deep models. Our experiments on vision datasets show that training feature attribution yields fine-grained, test-specific explanations: it identifies harmful examples that drive misclassifications and reveals spurious correlations, such as patch-based shortcuts, that conventional attribution methods fail to expose. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_09135 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Training Feature Attribution for Vision Models Bacha, Aziz George, Thomas Computer Vision and Pattern Recognition Machine Learning Deep neural networks are often considered opaque systems, prompting the need for explainability methods to improve trust and accountability. Existing approaches typically attribute test-time predictions either to input features (e.g., pixels in an image) or to influential training examples. We argue that both perspectives should be studied jointly. This work explores *training feature attribution*, which links test predictions to specific regions of specific training images and thereby provides new insights into the inner workings of deep models. Our experiments on vision datasets show that training feature attribution yields fine-grained, test-specific explanations: it identifies harmful examples that drive misclassifications and reveals spurious correlations, such as patch-based shortcuts, that conventional attribution methods fail to expose. |
| title | Training Feature Attribution for Vision Models |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2510.09135 |